import numpy as np
from scipy.stats import kurtosis, skew
import pandas as pd
import math
import matplotlib.pyplot as plt
import seaborn as sns

df = pd.read_csv('aug 4 2018.csv', header=None)
df.columns = ['seq1', 'start1', 'end1', 'res1', 'chain1', 'pdb1', 'seq2', 'start2', 'end2', 'res2', 'chain2', 'pdb2', 'rms', 'occ' , 'count']

df.head()

	seq1	start1	end1	res1	chain1	pdb1	seq2	start2	end2	res2	chain2	pdb2	rms	occ	count
0	ALAASNLYS	683	685	1.94	A	4Y03	ALAASNLYS	362	364	3.20	A	4Y3C	2.83927	111	3
1	ALAASNLYS	683	685	1.94	A	4Y03	ALAASNLYS	444	446	1.66	A	4Y04	3.59824	111	3
2	ALAASNLYS	683	685	1.94	A	4Y03	ALAASNLYS	362	364	3.20	B	4Y3C	2.10710	111	3
3	ALAASNLYS	683	685	1.94	A	4Y03	ALAASNLYS	73	75	2.04	A	4Y4V	2.59099	111	3
4	ALAASNLYS	683	685	1.94	A	4Y03	ALAASNLYS	362	364	3.20	C	4Y3C	2.99060	111	3

id_rms_df = df.groupby(['seq1'])['rms'].apply(list).to_frame()
id_rms_df = id_rms_df.reset_index()

num_bins = np.arange(0, 3.1, 0.1)
# bins
x_x = np.arange(0, 3.1, 0.1)
y_y = np.arange(0, 30, 10)
# y = [0, 10, 20, 30, 40, 50, 60]

# example data
for index, row in id_rms_df.iterrows():
    fig, ax = plt.subplots(figsize=(10,4))
    
    # the histogram of the data
    n, bins, patches = ax.hist(row['rms'], num_bins, density=1, edgecolor='black', linewidth=1.2)

    bin_centers = 0.5 * (bins[1:] + bins[:-1])
    ax.plot(bin_centers, n, color='red') ## using bin_centers rather than edges

    ax.set_xlabel(' -- RMSD -- ')
    ax.set_ylabel(' -- Frequency -- ')
    ax.set_title('RMS Distribution of ' + row['seq1'])
    ax.text(2.3, 16, 'Sequence: ' + row['seq1'], style='italic',
        bbox={'facecolor':'red', 'alpha': 0.5, 'pad':10})
    ax.text(2.3, 13, 'Skew: ' + str(skew(row['rms']))[:8], style='italic',
        bbox={'facecolor':'yellow', 'alpha': 0.5, 'pad':10})
    ax.text(2.3, 10, 'Kurtosis: ' + str(kurtosis(row['rms']))[:8], style='italic',
        bbox={'facecolor':'lightblue', 'alpha': 0.5, 'pad':10})
    ax.set_xticks(x_x)
    ax.set_yticks(y_y)
    # Tweak spacing to prevent clipping of ylabel
    fig.tight_layout()
    plt.show()